tubes sound

[planet10]

do you think that an event at some time can affect an event that happened previously? E.g., could something I do today affect my history?

Given that we (most of anyway) are on a 1 direction thru time, we would never know. At least in an absolute sense -- if it exists. Of course history is always changing as it is processed thru the human mind.

dave

 

[rdf]

It's income tax season in Canada. My history is revised daily.

 

[ray_moth]

Even if the diodes exhibit no reverse recovery effects, "hash" can be created by as the diodes abruptly turn off by the stray inductances and capacitances in the circuit.

I am led to believe that, if one is forced to use a bridge rectifier in a hi-fi amp, by virtue of having no centre-tap in the power transformer secondary, then a Graetz hybrid bridge, using TV damper diodes in combination with (decent) SS diodes, is as close as you can get to the best of all worlds. You get soft start with half a minute's delay, very little voltage drop (low sag) and no noise at all from the SS diodes. Is this true, do you think?

 

[jlsem]

Music is not very periodic... Fourier analysis can tell us some things but not all....

Which explains why push-pull amps do not necessarily have less distortion than do single-ended ones and feedback doesn't really reduce distortion either, but adds it....

John

 

[SY]

Given that we (most of anyway) are on a 1 direction thru time, we would never know. At least in an absolute sense -- if it exists. Of course history is always changing as it is processed thru the human mind.

dave

Well, since I don't understand that (sorry, I'm rather simple), let me ask for a concrete example. If I drop a ball off a rooftop on an airless planet, the tenth one perhaps, and the height of the rooftop is known, I can calculate to an arbitrary precision how long it will take for that ball to hit the ground using basic equations of gravitation. Do I need to account in that calculation for the possibility that the same experiment repeated the next day might be altered because, several hours after performing the first experiment, the terraforming project I was running removed 25% of the planet's mass?

Sorry to be a a pain, I'm just trying to understand your notions of non-periodicity.

 

[anatech]

I think people who gamble at casinos may have a problem with the answer to this.

-Chris

 

[poobah]

SY,

In the case of the tenth planet... you had better not use too large of a ball...

 

[planet10]

let me ask for a concrete example. If I drop a ball off a rooftop on an airless planet,

In a simple case like that results can be easily predictied. No infinities. And simple. What happens to your experiment if you toss a VERY large number of VERY small objects off at the same time (actually in practise you'd never be able to release them all simultaneously)

In the case of music, enuff complexity has creeped in that complexity/chaos theory starts to play a role. And since we cannot actual replicate an infinite process a real Fourier Transform is purely an intellectual exercise.

And in the case of the Discreet Fourier Transform that everyone actually uses, there is also a sort of "heisenburg principle".

What i have issue with is that because the tool yields some useful information, they start to believe that it is all the information (not explicitly said here anywhere, but implicit)... if the only tool you have is a hammer, you treat everything like a nail.

dave

 

[poobah]

SY,

Slap an electric guitar on an o'scope... Fourier doesn't break down per se... but it is not telling the whole picture.

 

[darkfenriz]

Yeah Fourier...
he was a clever guy
nice discussion Dave, but 'Fourier analysis' knows its own limitations, Heisenberg principe or Gibbs' effect being most spectacular.
Here we come damn close to philosophy- if you execute a dft program today and tommorow will you create a spectogram?

 

[SY]

But Dave, music doesn't have an infinite period, even if you include the time it takes for the reverberation to die down below thermal limit. In my gedankenexperiment, the realism of the instruments' sound is not, unless I'm missing something, depend on what the musician does when the playing is finished.

 

[lndm]

Hi SY,

It could be that you've demonstrated that CD's hold all the info we need and we should consider them perfect! (no, don't really want to talk about that).

I have always assumed that fourier transforms give enough info to put us on the right path. A path that will lead us where we want to go.

If I had any more detail, I wouldn't know what to do with it.

 

[planet10]

But Dave, music doesn't have an infinite period

But the Fourier Series to completely describe it is....

Attached is an analogy of the sinewave with the same periodicity of a typical piece of music.

dave

 

[SY]

It's only infinite if you don't do any bandwidth-limiting. Would you accept that a gigahertz acoustic bandwidth limit for a real acoustic event would be well beyond the bounds of audibility?

 

[lndm]

tubelab.com,

Would you agree that the mercury vapour rectifier would make a poor guitar amp rectifier? Would you consider one for hifi use?

 

[planet10]

It's only infinite if you don't do any bandwidth-limiting. Would you accept that a gigahertz acoustic bandwidth limit for a real acoustic event would be well beyond the bounds of audibility?

No, the fourier transform is infinite as it is a theoretical construct. The discrete fourier transform (ie approximation of the fourir transform) is not if band-limited (althou in the real world it is still not possible to completely satisfy the required initial conditions because no filter will satisfy the band limitations-- so we end up with a contaminated approximation)

dave

 

[arnoldc]

Bingo. Or fix your circuit.

Guys, how do I explain this-

I started my 12B4 project with 6X4 rectifier CLCLC, I liked it a lot but I believe some noise is coming from the 6X4. I changed it to 5Y3 and it's more silent.

Then I changed it so FR107 bridge config, with CRCRCRC filtering and seems to be more silent.

Then I grafted a 200V PSU with IRF840 and I can't believe what I heard- greater dynamics and slam, better resolution, layering and soundstage, coherence and control.

I placed a scope and watched the B+ and it seems to be "stable" with whatever I feed (music) into the pre.

I placed the CRCRCRC back and watched the scope going haywire, nasty spikes, distorted line (sometime becoming an S shape), sags too.

Never saw what the 6X4 looked like but now I won't go back to simple tube rectifers with CLCLC.

Maybe bad tube psu design? or am I comparing apples and oranges?

 

[Miles Prower]

Would you agree that the mercury vapour rectifier would make a poor guitar amp rectifier?

If what you want is a PS with poor regulation, then a MV diode is most definitely what you don't want. Like a silicon diode, the MV diode has a forward voltage that doesn't change very much with load current. This is why you see them in ham rigs: you don't want the DC to bounce up and down when you key the xmtr. Makes for poor sounding CW notes.

Would you consider one for hifi use?

No, I'd use silicon. Less forward voltage drop, no heater, no worries about mercury spills, one less socket. All the advantages of MV with none of the drawbacks. About the only reason to opt for MV over silicon is if you just have to have those glowy bottles.

 

[Larry Lomax]

Hello folks,
I enjoy this discussion a lot. And I feell I can reply with my own opinion.
I was curious about what is music by the point o view of the electric signal is re-produced by a media, CD, vinyle records, tapes, or even pickups from guitars etc...
Well, mi scope tells that in the time domain (X axis) voltage, and thus current on a pure resistive load, is a garbled wave (you'll need a memory scope to see it) composed by the sum of all the single waves produced by the music in the moment the sample is recorded. This wave can be further analysed using FFT, and check the spectrum of the power components of that garbled wave in the frequency domain rather than in the time domain. Well, all of this verify that Fourier was right. All we have to consider is that in the audio band, as to say, all musicall instruments and uman voice and audio noise are limeted by about a 50Khz upper limit. So, this is the limit for the fourier analisys. In other words, the rising time of the fastest (and furious) wave, will not produce allmost any power over 50-60Khz and shurely not over 100Khz. This is not a limitation of the fourier analisys but the limitation of the audio system dictated by a further limitation of the uman hear.
Please, leave the Heisembergh indetermination to the quantum phisics, where it pertains. 😉
Regards.

 

[lndm]

There is the concern that standing waves can be activated by a persistent wave, but that a transient portion f a wave will not cause such a build up, and then that room reflections are not so discriminatory.